Do you find Pisa still available during the later game, when you are approaching the time for your GS bulb-fest? The AI doesn't usually beeline Pisa, like it does some other wonders (mainly because Pisa is on a direct line to a few civs', but not that many civs', unique units, which is usually the driver for early AI wonder-building), but I rarely see it lasting that late in my games (and not just because I'm the one who builds it earlier

).
But you raise an interesting question: is getting an early GS and +25% GPP in all cities actually a bad deal for GS generation? Since I can't recall ever mapping out the arithmetic, I decided to noodle through it this evening.
A Pisa GS (or GE or GM, for that matter, and the GS from PT) will increase the cost of every subsequent GS by 100 GPP. So, if you built Pisa when you had already generated 2 GSs, the Pisa GS would come in place of your 300 GPP GS, so your next GS would be 400 GPP, then 500, 600, etc.
Let's assume your goal was just to generate 8 GSs from GPP sources, without the distraction of Hubble, Porcelain Tower, Pisa, the ISS or the Liberty finisher. In that case, you would need 3,600 cumulative GS GPP to spawn 8 GSs.
So let's explore two scenarios. In each scenario you goal is 8 naturally spawned GSs (could be 9, 10 or more, but I don't think that affects the analysis), plus 1 from Pisa, 1 from PT, 2 from Hubble and 3 from faith (the 1,000, 1,500 and 2,500 faith GSs, for total 5,000 faith). This means you aren't going Order (for the Tier 3 tenet that provides 1 GE and 1 GS), using the Liberty finisher for a GS, or building the ISS, but I think it is all OK for purposes of this analysis.
I'm also ignoring the effects of Humanism (+25% spawn rate for GSs) and any tenet that boosts Great Person generation rate -- both of those would speed up the turn counts for GS spawns, but you can take those, or not take those, whether you built Pisa early, built it late or never built it at all. They are easily enough layered into the computations, but I chose to leave them aside, just to keep this as simple as possible.
So, in each scenario you have a 4-city Tradition empire, with a Garden in each city and National Epic in your capital. You get Education on turn 100 (just to simplify the turn count, but that is also a reasonable turn time, though perhaps a bit on the speedy side for many civs), Scientific Theory on turn 150, and Plastics on turn 200 (again, simplifying the turn counts, but those are not far off a reasonably speedy pace). In each case, you have enough gold to rush-buy 4 universities, 4 public schools and 4 labs on the turns you finish those techs, and you work those scientist slots immediately and continually. And finally, to again simplify the turn count, we assume you want to begin your GS bulb-fest on turn 230, just after you finish Hubble and buy the last of your faith-bought GSs.
In scenario 1, you are confident that you can delay building both Pisa and PT until after you have generated your 8 natural GSs, so the sequence in Scenario 1 is generate 8 GSs, build Pisa, build PT, build Hubble, faith-buy 3 GSs, and then bulb -- all by turn 230. In scenario 2, you slip Pisa into the build order to finish after you generate natural GS 2 and before you generate natural GS 3. PT and Hubble still get built after you've generated 8 natural GSs, plus the "free" GS from Pisa. And in each scenario, we assume you generate no GEs or GMs (ever), other than a single faith-bought GE to rush Hubble.
In both scenarios, you would expect to generate your first GS (from your capital) on about turn 112 (Note: for all of this, I'm ignoring GS points from wonders, like Oracle -- I don't think they affect the analysis, but they are an arithmetic complication I am ignoring). Your second GS then pops out of city 2 on about turn 127.
In Scenario 1, you still need to generate another 3,300 GS points to spawn 6 more GSs. They should pop out as follows:
GS 3 - about turn 140 in city 3
GS 4 - about turn 144 in your capital
GS 5 - about turn 162 in city 4 (public school on turn 150 speeds it up a bit)
GS 6 - about turn 170 in city 2
GS 7 - about turn 179 in city 3
GS 8 - about turn 205 in your capital
Note that, if we didn't build Pisa, PT and Hubble after GS 8, your natural GS 9 would spawn in city 4 on about turn 231. But if you generate 4 GSs from those wonders, GS 9 will be greatly delayed (in this scenario it would be generated in your capital somewhere around turn 270, depending on when you actually built Pisa). So, depending on how fast you expect to finish that game, delaying Pisa and PT until after turn 231 (in this example) might be helpful.
Anyway, in scenario 2, you build Pisa between GS 2 and GS 3 -- let's say that Pisa is completed on turn 130, for ease of computation. GS 3 is now actually GS 4, and costs 400 GS points, instead of 300, and so forth through natural GS 8 (really GS 9). So, instead of your natural GSs costing you a cumulative 3,600 GS points, they cost a cumulative 4,200 GS points (6 GSs for an extra 100 GS points apiece). But Pisa does provide an additional 25% GPP from the time it is built, so how does all that affect GS spawn timing?
Scenario 2's GSs 3 through 8 (really, GSs 4 through 9, since Pisa's is really GS 3) should pop out as follows:
GS 4 - about turn 143 in your capital
GS 5 - about turn 158 in city 3
GS 6 - about turn 169 in city 4
GS 7 - about turn 179 in city 2
GS 8 - about turn 194 in your capital
GS 9 - about turn 218 in city 3
Yes, it does appear that taking a Pisa GS will slow down the spawning of subsequent GSs by a bit.
So,
IF you are confident that you can delay taking Pisa that long (I'm not, but YMMV), you would benefit by delaying building Pisa until you are done spawning natural GSs.
Of course, if you lost Pisa to another player, you would need to generate a 9th natural GS to fill the gap left by the missing Pisa GS, which would force you to delay taking PT and Hubble until after GS 9 spawns (around turn 230 in Scenario 1).
Anyway, this was very interesting. Thanks for raising this point.